The Mission: Building Pathways to Upward Mobility
At Transfr, the goal was to build a direct bridge from the classroom to a career. As a Senior Technical XR Designer, I've led the design and development of 25+ VR simulations across technical trades, from automotive technicians to pipefitting specialists. They're used in high school classrooms and workplace facilities across the U.S., by 100,000+ users.
My role on each simulation:
- Build the scene in Unity and the simulation logic on the Transfr SDK, working with instructional designers, technical artists and sound designers.
- Work with SDK developers to improve the tools we build with: interaction templates, nodes and components, voice-over generation and NPC animations.
- Solve the interaction problem unique to each career, so a physical task feels natural in VR.
Deep Dive: The Flagger Sim
Flagger trains learners to direct traffic through a road work zone: holding a STOP/SLOW paddle in one hand, giving slow, stop and go hand signals with the other, and coordinating with a second flagger over the radio. In September 2025, E2Engine built Flagger's initial structure and saved ~4 days of setup. I put that time into the sim's three hardest design problems.
Design Challenge 1: Hand Signals From Controller Movement
Hand signals are the main interaction throughout the sim — they're how the learner controls traffic. The challenge was reading a hand gesture with nothing but a hand controller: no hand tracking and no buttons to press for each signal.
- Orientation and movement: the sim reads the controller's orientation and how it moves to work out which gesture the hand is making — stop, slow or go.
- Hand poses follow detection: as a signal is detected, the virtual hand changes to that signal's pose, so the learner sees what the sim has read and can correct it.
- Color for clarity: a highlight shader outlines the hand in the color of the detected signal — red for stop, yellow for slow, green for go — so the signal reads at a glance.
- Learn, then perform: the lead flagger demonstrates each signal and the learner copies it, then uses the signals to control live traffic for the rest of the sim.
Design Challenge 2: Choosing Where to Stand
In one part of the sim, the flagger has to decide where to stand, and the choice itself is what's being taught. A normal teleport would let the learner wander anywhere, so we designed a multi-select teleport: the learner picks between two spots on a minimap of the street.
- Default: a small flagger figure stands beside a minimap with two pins, and the prompt reads "Pick up and place to teleport".
- Grabbed: the learner picks up the figure and places it on one of the two spots.
- Placed: a radial countdown runs before the teleport, and the learner can grab the figure again to change their mind.
- Character: the sketch gave the figure personality — it reaches up on hover, goes limp like a ragdoll when grabbed, and waves goodbye as the countdown finishes.
Design Challenge 3: Showing the Consequences
The mistakes that matter most for a flagger are the ones that cause traffic accidents — exactly what can't be practiced safely on a real road. VR can show them safely, so we decided to simulate the real consequences instead of just marking an answer wrong.
- Every action has an outcome: a wrong hand signal, or forgetting to radio the other flagger, plays out as a traffic consequence inside the sim.
- Learn from the result: the learner sees what their mistake would cause on the road, then gets the chance to do it right.
- Assessed too: wrong signals are also recorded for assessment, so instructors can see which ones a learner confused.
Bridging Design, Product, and Engineering
We were building the car and the factory at the same time. Alongside the simulations, I worked with multi-disciplinary teams on Transfr's Unity SDK and its XR interaction system, as the company moved from single prototypes to a library of simulations.
- Interaction patterns: I helped turn real-world physical tasks — operating a CNC machine, handling laboratory tools — into standard XR interaction patterns. The tool templates became the VR Design System for Hands-On Tools.
- SDK tools: I helped shape Node Flow Studio and Asset Manager, which keep visual and instructional standards consistent across modules.
- Production workflows: I designed AI-driven workflows to speed up simulation development, which grew into E2Engine and Scriptr.
- Selected simulations (videos below, in order): knee surgery assist, automotive oil change, hospital power outage, residential solar panel install, commercial pipe preparation and diesel mechanics.
Real-World Impact
These are company-level results for the library my simulations are part of:
- Learning gains: Transfr's efficacy studies found its simulations often produced higher learning gains than real-world practice alone. Read the Coastal Bend study: transfrinc.com ↗
- Reach: 100,000+ users across the U.S., in high schools and workforce programs.
- Learner feedback: user testing with learners fed back into both the simulations and the SDK tools.